transformers/examples/pytorch/language-modeling/run_clm.py
Simple question, I can’t figure it out myself. This is a nice utility script for training a model, but the output is pretty verbose, and ugly:
When I’d trained models before, I got a super nice clean, semi-graphical output with a nice grid of evals and epochs:
is there anyway to turn this on for the run_clm script please?
Thanks in advance,
W
Incidentally, these are both being run inside colab. I suspect that this might be caused by the way that I’m running the run_clm script, as a !python run_clm command. Whereas the nicely formatted it one is being run as a trainer.train() evocation in the colab cell itself.
If there’s a neat way to turn the run_clm script into a simple native colab python execution, that might work, I just wanted to avoid the hassle of cut and pasting very large amounts of parameterized Python into my colab. Eventually I probably will, but right now, I wanted to just have a quick and pretty runtime experience while I’m playing with distilGPT2 training.
